Approccio meccanico statistico allo studio delle Macchine di Boltzmann Ristrette

La Macchina di Boltzmann Ristretta (RBM) è una rete neurale a due strati utilizzata principalmente nell'apprendimento non supervisionato. La sua capacità nel rappresentare complesse relazioni insite nei dati attraverso distribuzioni di tipo Boltzmann Gibbs la rende un oggetto particolarmente interessante per un approfondimento teoretico in ambito fisico matematico. In questa tesi vengono presentati due ambiti di applicazione della meccanica statistica all'apprendimento automatico. 1) La similarità della RBM a unità binarie con il modello di Ising permette di sfruttare un'espansione alle alte temperature per approssimare l'energia libera, termine presente nel gradiente della likelihoood e difficile da trattare numericamente. I risultati ottenuti con questa tecnica sul dataset MNIST sono paragonabili a quelli ottenuti dalla Contrastive Divergence, che utilizza invece metodi di Monte Carlo. 2) L'equivalenza statistica della variante ibrida di RBM con il modello di Hopfield permette di studiare la taglia del training set necessaria per l'apprendimento attraverso l'analisi del problema inverso, in cui i ruoli di spin e pattern sono invertiti. Viene quindi presentato un metodo basato sulla teoria di Gauge che permette di derivare il diagramma di fase del modello di Hopfield duale sulla linea di Nishimori in funzione della temperatura e del rapporto tra numero di campioni e dimensione del sistema.

Transizioni di fase nell'allenamento delle Macchine di Boltzmann Ristrette

In this thesis, we dealt with Restricted Boltzmann Machines with binary priors as models of unsupervised learning, analyzing the role of the number of hidden neurons on the amount of examples needed for a successful training. We simulated a teacher-student scenario and calculated the efficiency of the machine under the assumption of replica symmetry to study the location of the critical threshold beyond which learning begins. Our results confirm the conjecture that, in the absence of correlation between the weights of the data-generating machine, the critical threshold does not depend on the number of hidden units (as long as it is finite) and thus on the complexity of the data. Instead, the presence of correlation significantly reduces the amount of examples needed for training. We have shown that this effect becomes more pronounced as the number of hidden units increases. The entire analysis is supported by numerical simulations that corroborate the results.

The “FRA wheel” as a teaching tool to elaborate on the identity aspects of physics and mathematics as disciplines in interdisciplinary contexts

This thesis project is framed in the research field of Physics Education and aims to contribute to the reflection on the importance of disciplinary identities in addressing interdisciplinarity through the lens of the Nature of Science (NOS). In particular, the study focuses on the module on the parabola and parabolic motion, which was designed within the EU project IDENTITIES. The project aims to design modules to innovate pre-service teacher education according to contemporary challenges, focusing on interdisciplinarity in curricular and STEM topics (especially between physics, mathematics and computer science). The modules are designed according to a model of disciplines and interdisciplinarity that the project IDENTITIES has been elaborating on two main theoretical frameworks: the Family Resemblance Approach (FRA), reconceptualized for the Nature of science (Erduran & Dagher, 2014), and the boundary crossing and boundary objects framework by Akkerman and Bakker (2011). The main aim of the thesis is to explore the impact of this interdisciplinary model in the specific case of the implementation of the parabola and parabolic motion module in a context of preservice teacher education. To reach this purpose, we have analyzed some data collected during the implementation in order to investigate, in particular, the role of the FRA as a learning tool to: a) elaborate on the concept of “discipline”, within the broader problem to define interdisciplinarity; b) compare the epistemic core of physics and mathematics; c) develop epistemic skills and interdisciplinary competences in student-teachers. The analysis of the data led us to recognize three different roles played by the FRA: FRA as epistemological activator, FRA as scaffolding for reasoning and navigating (inhabiting) the complexity, and FRA as lens to investigate the relationship between physics and mathematics in the historical case.